Ryota Takatsuki

PhD student @SussexCentre. Research fellow @AIAlignNetwork. Dreaming of reverse-engineering consciousness someday.

[Ryota Takatsuki @rtakatsky]

🎓 Life update
This month, I’ll be moving from Tokyo to Brighton to begin my PhD program at @anilkseth's lab (Sussex Centre for Consciousness Science)!
Looking forward to further exploring the intersection of AI & consciousness as part of this amazing team!

[Koshiro AOKI @KoshiroAoki]

Can LLMs control their own 𝘱𝘳𝘪𝘷𝘢𝘵𝘦 representations?

To test this, we propose in-context neurofeedback, inspired by neuroscience techniques. Our results suggest current LLMs do not yet have this ability, but we offer a more rigorous test of metacognition for future LLMs.

[Anil Seth @anilkseth]

1/3 Conscious AI and Biological Naturalism. A veeery loooong time in the making, but I'm happy to say that the full collection is now out in @BBSJournal 🔥 I deeply thank all commentators - of fifty commentaries! - for their wise, broad, and deep opinions on this critical topic. I also thank them for their saintly patience. 🙏🏽

I remain highly skeptical about the prospects for conscious AI (for silicon digital systems), but I've learned an enormous amount from the privilege of engaging with these commentaries. Hands down one of the most rewarding experiences of my professional life.

My hope is that - whether you agree with me or not - having such a diversity of views collected in one place will serve a useful resource for discussions about consciousness and AI.

The original target article is available here: https://t.co/Tsaslytu56

My response "The stuff matters: consciousness, computation, and biology" was published just now, and is here: https://t.co/gtOewuG9MY

And the set of 50 commentaries are collected here: https://t.co/HUhiraO3kG

Unfortunately these are not open access. I will try to remedy that, at least for the target & response.

FWIW there's a more accessible version of my view on the topic here in @NoemaMag https://t.co/LmSiQnT9Wh and also a short @TEDTalks from earlier this year https://t.co/vDvw82ookk

[Hitomi Yanaka (谷中 瞳) @verypluming]

#EMNLP2026 mainに2本論文が採択されました🎉トークンレベルの最適輸送による言語間アライメントでLLMの多言語能力を改善させた話と、双安定性錯視でLVLMと人間の知覚メカニズムを分析した話です！Our two papers have been accepted to #EMNLP2026 main! Stay tuned w our papers on arXiv! #NLProc

[Anil Seth @anilkseth]

1/2 My take on the latest @AnthropicAI claims about potential signatures of consciousness in Claude: https://t.co/ViSgTJydxr

[Ryota Takatsuki @rtakatsky]

Link to the paper: https://t.co/wBLGEAZUQk

[Ryota Takatsuki @rtakatsky]

Our work, led by @KoshiroAoki, will be presented as a spotlight at the ICML AI4GOOD workshop!

We revisit recent claims of metacognitive control in LLMs by introducing in-context neurofeedback—a stricter task where success requires privileged access to internal activations.

[Koshiro AOKI @KoshiroAoki]

I’ll be presenting two papers at #ICML2026 workshops (AI4GOOD WS & Mech Interp WS).
Let's talk if you're interested in introspection, metacognition, or differences between humans and LLMs!

[Anil Seth @anilkseth]

Today at @theASSC #ASSC2026 - Ryota Takatsuki @RyotaTakatsuki will present his poster (poster session 1): Do Autoregressive LLMs Exhibit Postdiction-Like Integration of Later Tokens? ... a very cool examination of whether LLMs have perception-like features .@SussexCentre

[Ryota Takatsuki @rtakatsky]

Presenting our poster at #ASSC29 today:

Do Autoregressive LLMs Exhibit Postdiction-Like Integration of Later Tokens? (P1-021)

Come say hi if you’re interested in time perception in LLMs.

Joint work with @Sho_Westbook, @GoukiMinegishi, and @anilkseth.

[Koshiro AOKI @KoshiroAoki]

論文が公開されました。こちらから読むことができます。
https://t.co/UZH6IR8n9b

[Anil Seth @anilkseth]

Timescapes of Animal Experience: new paper in @TrendsCognSci, led by Ishan Singhal @ishan276 w/ @birchlse & me. Open access link: https://t.co/whg0sW3boP .@SussexUni .@SussexCentre

[Ryota Takatsuki @rtakatsky]

人工知能学会40周年という節目に、我々のmechinterpサーベイが記念論文賞の最優秀論文賞を受賞しました！
近年急速に進展している本分野を日本語で網羅的に整理しているので、ぜひ読んでみて下さい！

[Koshiro AOKI @KoshiroAoki]

Mechanistic Interpretability（機械論的解釈可能性）に関するサーベイ論文が、人工知能学会40周年記念論文賞 最優秀論文賞に選ばれました。人工知能学会論文誌「人工知能学会40周年記念特集号」（2026年7月号）に掲載される予定ですのでぜひご覧ください。

[Ma Shuqin @Shuqin_Ma]

New preprint with @kanair. https://t.co/9CPpjbOPBK
Anti-computational arguments are right that externally assigned labels can't ground consciousness. But labelling and intervention-based individuation are not the same operation.

[Anil Seth @anilkseth]

Fully funded PhD position at @SussexUni at the @SussexCentre for Consciousness Science. Stipend plus fees for 3.5 years, kindly supported by the School of Engineering & informatics. Apply by May 31st. More details here: https://t.co/U3atEuAoYs

[Gouki Minegishi @GoukiMinegishi]

Our paper was accepted as a #ICML2026 Spotlight!

Reasoning in LLMs has improved largely by chaining local steps. But is that the whole story?
Humans occasionally make inferential "leaps" across domains, a faculty known as analogy.
We design a synthetic task to show how small Transformers acquire analogical reasoning, and find that the same signatures appear in pretrained LLMs.

arxiv: https://t.co/1WCizIKWly
code: https://t.co/82kOKCtJo7

[Ryota Kanai @kanair]

I’m happy to share our new preprint on “canonical functionalism.” (with @Shuqin_Ma)

The core idea is that functional structure can be uniquely identified by the system’s counterfactual state-transition structure, without arbitrary semantic labels.
https://t.co/WwfrtVq0Tg

[Anil Seth @anilkseth]

1/2 Why AI is unlikely to become conscious – my 2026 @TEDTalks is now online. What do you think about the prospects for 'conscious AI'? https://t.co/OoXFavR3eG

[Perceptual Timeline in AI (AI時間構造学) @COlympia_PTA]

3/6に開催された、academist Prize の中間報告会にて、AI時間構造学が「アカデミスト賞」を受賞し、賞金30万円をいただくことができました！　こうして研究が続けられているのも、皆��まのサポートがあってこそです。引き続きよろしくお願いいたします。
#academist

[⿻ Yusuke Hayashi 林祐輔 @hayashiyus]

【コワーキングスペース設立に向けたご協力のお願い】
AI安全性，AIアライメントの技術的な研究チームを東京で組成しよう，という話が進んでいます。そのコワーキングスペースの設立に向けて，ぜひ皆様にEOI（関心表明）のご提出をお願いしたいです。10分程度で終わる簡単なフォームです。
https://t.co/n5HfohRr84

現在，ロンドンのAI Safetyスペース「LISA」の創設者が，アジアでの新拠点として東京を検討しています。東京が候補地として選ばれるためには，皆様の関心の高さを示すことが非常に重要です。
みなさまご指摘の通り，コミュニティの形成と成熟には物理的な拠点が欠かせません。東京での実現に向けて，ぜひご協力をお願いいたします。

[Anil Seth @anilkseth]

1/ I am truly honoured and completely delighted to have won the 2025 Berggruen Essay Prize (English), for my submission “The Mythology of Conscious AI” @berggruenInst @NoemaMag https://t.co/n3eXSRXkvk

[Anil Seth @anilkseth]

1/🧵 Granger Causality Maps for Langevin Systems – new preprint from the incomparable Lionel Barnett, w/ Benjamin Wahl, Nadine Spychala, and me – a @SussexCentre for Consciousness Science production. Link at end of thread.